RidiculousHat is super active on Reddit. I wonder if someone asks the question there if he would respond.
I asked a question last week about interaction with Tolin’s Goblet and he responded confirming there was a copy bug and it would be fixed.
- That infographic is from before 2020, so it applies to a Ranked system that no longer exists.
- Gaming journalists are lazy and will repeat point #1 because that’s the first thing Google tells them. I believe that almost all news is fake, but I do NOT believe that this is caused primarily by malice; I believe it’s caused by low standards and apathy towards standards. Repeating #1 (only 0.5% of players make Legend) in the present day is a great example.
- If we define a Hearthstone player as anyone who has at any time created a Hearthstone account, then approximately 4% to 4.5% of players make Legend each month. However, I don’t wish to include “tourists” in my numbers, and in any free to play game the vast majority of players are tourists. I am openly saying that about 7/8ths of the Hearthstone accounts don’t even count, because you can count the total number of games they’ve played this month on one hand, therefore they aren’t really Hearthstone players at all. That’s why I said regular Hearthstone players, and not just Hearthstone players.
Now if you want to say that #3 is some form of “no true Scotsman” fallacy, maybe it is. I am certainly doing something very close to it. But I stand by it, because any analysis of a free to play game is going to have the vast majority every month being people who installed for the first time only to permanently uninstall within the exact same month. There is no way that this larger group of players is representative of the smaller group of dedicated players who actually sustain the game in any meaningful capacity.
1 Like
This is a fair way of interpreting what I’m saying. Well, I said 35%, but other than that.
1 Like
Not it’s fine (not a true scotsman). The game is filled to the brim with people who don’t take Standard seriously; hell even Wild players dilute the results; I dilute the results of Wild by doing achievements there and going up to Platinum or Gold.
It’s reasonable to assume that only 1/3rd of those taking it seriously to go Legend achieve it; the only real subtlety is that we have very different ways to even do that; e.g. I play extremely few games because I try to maximize efficiency.
Ummm……. us mobile players arent serious? 
You’re making this number up.
And making this one up
You’re also completely making up what the number of “regular players” are.
We don’t have any numbers, you’re just guessing but you’re not admitting your guessing could be way off.
1 Like
Yeah…
I’m a regular legend player, and probably 98% of my games played have been on mobile, so no tracking for those sites.
The only time I run a deck tracker is for getting replays.
3 Likes
Thanks, you guys are further proving my point. I don’t think he realizes how many people DON’T run the app. Like almost none of my friends on my friends list do.
1 Like
Not really. I’m taking the 0.5% prior to the 2020 rank change, and applying a ~8x multiplier. Evidence for the ~8x multiplier is here: https://hearthstone-decks.net/how-many-players-reach-legend-in-hearthstone/ scroll down to April 2020.
I see that you found an infographic that says 0.3%, I’ve seen a very similar infographic that said 0.5%. I wasn’t aware of the 0.3% version.
What I’m saying is that players who have deck tracker software installed are roughly representative of non-tourist, “real” players. (Please note, Mogkuppo, that representative of doesn’t mean exactly the same as.) I have already identified this as the largest leap in my logic.
I am just guessing, and my guessing could be way off. I am, however, trying to guess to the best of my ability. That’s why my guess has data citations.
Most Hearthstone players play the game on mobile, and therefore don’t run deck trackers. I don’t run deck trackers. But you don’t need to sample a majority of a population, or even anything close to a majority, to get a representative sample.
No, but what you’re doing isn’t the same as sampling a part of a population to get a feel for its entirety. The total real players without a deck tracker are unknown. You’re sampling only the ones running the tracker and guessing the bigger population.
That would be like sampling all the people that live in 1 state and guessing how many people live in the entire US based on that 1 state.
Without knowing how many “real” players don’t run the app, you are going purely off speculation and that number must be known to be able to guess accurately.
Knowing the population of the US is ~330 million and sampling 3 million is vastly different from not knowing the population of “real players without deck tracking” and just sampling those with it because in this case, those people with trackers are almost all really serious about playing the game whereas those without the tracker can be a mixture of both. It would be like polling Conservatives only to find out how many Liberals and Conservatives altogether feel about a single Conservative issue.
1 Like
If you’d like, I think it’s fair to express my claim as an if-then statement.
IF players who have deck tracker software installed are representative of “real” non-tourist Hearthstone players, THEN finishing the month around Diamond 5 is average for those players AND finishing in Legend puts one in approximately the top third of those players.
I think it’s reasonable to believe in the IF here, but I readily confess that it’s far from proven.
I think I’ve been reasonably ethical about implying this IF from the very beginning. I mentioned deck tracker software from the first.
1 Like
Ok, but let me provide you with another IF THEN statement
IF a player doesn’t have a deck tracker software installed, THEN they are not a real “non-tourist” player.
That’s the problem in the logic. We know that is provably false. And it’s not just a little off, it’s way off.
The best statement you can make is that 30% of people who install a deck tracker get to legend. That’s it.
1 Like
I was never implying this. It is possible for a sample to be representative of individuals who are not in the sample. If this was not possible, then all polling would be invalid.
I get what you’re trying to say, but let me put this in a way I believe we can both agree on.
You’re taking the sample size of deck tracker users and applying that to a pool of “all serious players” because deck tracker users are a sample of “all serious players”.
I get that.
But here’s where I think the fault lies.
Let’s do another sample size.
Let’s only take the sample size of NFL quarterbacks.
30% of NFL QB can throw at least 70 yards.
NFL quarterbacks are a sample size of all quarterbacks that exist in the US (that includes high school quarterbacks)
Thus, 30% of all quarterback in the US that exist can throw at least 70 yards.
But in reality, only 3% of quarterbacks in the US can throw at least 70 yards.
This is where I think the breakdown is.
All QBs in the NFL are serious
All people with deck trackers are serious
Some QBs in the US are serious, but not all
Some people without deck trackers are serious, but not all
You can’t take a sample size of something that is different from the total you are trying to apply the data to.
You can’t take data just from Group A and apply it to Group A and B as a total if Group A and B are different
You must take the sample size from both Group A and B equally
1 Like
I don’t like your analogy at all. The reason is that the NFL is a highly selective, meritocratic organization. In contrast, anyone can download deck tracker software.
A better version of your analogy would be if we tracked “serious” quarterbacks by whether or not they’ve made purchases of (American) football equipment at one of a list of major sporting goods stores. A sporting goods store is not selective about throwing skill; they’ll sell to pretty much anyone.
It’s possible that this filter would exclude multiple NFL quarterbacks, who have never bought their own equipment in their lifetimes, instead having it provided by their teams.
All that said, I am still aware that the degree to which such a sample would be representative is a hypothetical thing.
The point wasn’t to compare selective with self selective.
The point was to compare great players with a mix of great and non-great players.
And then show why you can’t apply great player polls and apply to the mix of great and non-great player polls.
It had nothing to do with being a direct analogy of a selective process.
Thus, the whole A and B thing.
A will skew A+B, especially when you don’t even know what B is.
For all we know, A players reach Legend 30% of the time and B players reach it 1%.
The most I can give is if you said “If you download a tracking app, you have a 30% chance to reach Legend” and I’m ok with that.
1 Like
Thirty-five percent.
Or whatever the tracking data supports, sure. I have no qualms with that.
1 Like
And if you are okay with that, then you’re also okay with “IF players who download a tracking app are representative of Brazilian martial artists, THEN Brazilian martial artists have a 35% chance to reach Legend.” I imagine that you would evaluate the IF statement as false, as would I, but if at some point later evidence came up which, to my great surprise and mild alarm, proved the relationship to exist, then I would accept the THEN part.
Of course, we can replace “Brazilian martial artists” with anything whatsoever.
Well of course I’d give it a false, but that’s because my statement about what I’d be willing to give was meant as a joke so you can post that on the forums telling people to download the app to increase their odds lol
Obviously, downloading the app doesn’t increase your odds. But telling people the data supports it IS the joke “because 35% of people who run deck trackers reach legend”.
If 100 new people downloaded the app, how many would get to Legend on average? 